The Biden administration is urging cloud service providers, including Amazon, Google, and Microsoft, to identify foreign users involved in creating AI technologies, particularly targeting users in China. A new proposal could dictate these companies to disclose the identities including IP addresses of international clients working on AI projects. This rule aims at limiting access to the essential data centers and servers needed for AI development and signifies increasing tensions between the American and Chinese AI industries.
Microsoft’s Vice President of Government Affairs, Fred Humphries, conveyed that the organization is preparing to put the draft rule into action, stating “We welcome know-your-customer and cybersecurity requirements for AI infrastructure and look forward to providing input on the draft rule.”
The AI industry of China, initially slow in adopting generative AI technologies, adjusted its course in mid to late 2023 when technology powerhouses such as Baidu and Tencent launched their own public offerings. Baidu purports to have developed a language model on the same level as OpenAI’s GPT-4.
In addition, the Chinese government has been vigilant to ensure that these offerings reflect the nation’s socialist values and avoid criticizing the government or its leader, Xi Jinping. Baidu’s ERNIE Bot was found to heavily censor political topics.
The US has responded to China’s advancements in AI technology by implementing various strategies to curb the growth of China’s AI industry, mainly blocking the use of high-end hardware manufactured by firms such as Nvidia for training AI models.
The Biden administration, under the leadership of Secretary Gina Raimondo, has been focusing on curtailing AI security threats at a national level, stating “These models getting in the hands of non-state actors or people that aren’t our allies is very dangerous.”
Despite the visible tension, the U.S. and China have reportedly held undisclosed meetings on AI safety. However, the new rule, open to public feedback until April 29, embodies the U.S.’s broader approach to preserving its technological lead and protecting its national security interests.